The present disclosure relates to an image processing device and an image processing method, and particularly to an image processing device and an image processing method which reduce noise included in an image.
In shooting an image, there is known technology for obtaining an image with reduced noise by compositing successively-shot multiple images (frames). For example, in the case where, on an image to be processed (target image), multiple images (reference images) are composited, which are shot successively before or after the shooting of the target image and are aligned by motion estimation and motion compensation, images which are almost the same as each other are integrated in a time direction. Accordingly, the noise randomly included in each image is cancelled out, and hence is reduced as a result. Hereinafter, the noise reduction (NR) achieved by such a method is referred to as frame NR.
In the past, a local motion vector with respect to a target block which is set within a target image has been estimated, and using the estimated local motion vector, global motion which represents the change applied to the entire images, between two images has been calculated. The global motion normally represents motion and amount of motion of a background as a still image part of an image.
As technology using the global motion, there is disclosed, for example, technology involving separating a background still image part and a moving subject-of-shooting part within an image, generating a motion compensation image (MC image) using a local motion vector which corresponds to a global motion vector generated from a global motion, and compositing the motion compensation image and a target image (for example, refer to JP 2009-290827A). In this technology, the MC image is generated using applicatively the global motion vector and the local motion vector, and compositing processing is performed.
In this technology, for example, when the reliability of the local motion vector is high, the compositing processing taking the moving subject-of-shooting into account can be performed, by using the MC image determined from the local motion vector. Further, in the case where the reliability of the local motion vector is low, the local motion vector is not used and the MC image determined from the global motion vector is used, thereby avoiding the generation of an MC image that is indeterminate as to whether the image includes a moving subject-of-shooting or noise, and hence, stable compositing processing can be performed.